Although the prevalence of dementia increases with advancing age, and the disability, which accompanies dementia is far-reaching, few effective treatment programs exist. Capitalizing on our knowledge of physical frailty in older adults, behavioral and affective disturbance in AD, and the impact of these problems on caregivers' affective health, this project is designed to develop, implement, and evaluate an integrated model of medical, behavioral, and psychosocial intervention. Treatment is designed to reduce the physical, behavioral, and affective problems in patients with AD and to alleviate the depression and burden in their caregivers. Two hundred community residing AD patient-caregiver pairs will be randomly assigned to the active treatment or to treatment-as-usual (acute medical care, advice and support, and crisis intervention). The active intervention is designed to simulate clinical practice by providing a step-up approach to care for all AD patients and their caregivers. The specific goals of intervention are to reduce physical disability by improving patient mobility, maintaining adequate nutrition, decreasing adverse drug reactions, and decreasing falls and fractures; to reduce affective disability by identifying and treating depressive symptoms in patients and caregivers; to decrease patient behavior problems; and to decrease caregiver burden. All subjects are members of a community-based HMO and are enrolled in an ongoing Alzheimer's Disease Patient Registry that identifies and follows incident cases of AD. Subjects will be assessed upon enrollment in the investigation, and at six month intervals for a period of at least two years to evaluate their cognitive, physical, behavioral, and affective status. Health care utilization and costs will also be evaluated. It is hypothesized that AD patients in the active intervention will show significantly better outcome than controls, as deflected in better health, fitness, and mobility, less depression, and fewer behavior problems over the two year period. Further, it is hypothesized that caregivers in the active treatment will be less depressed and burdened than caregivers in the control condition. Finally, reduced health care utilization is expected in the active treatment group compared with the control group.